DICOM Image Normalization for Cross-Device Display
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of the DICOM standard for medical images limits their processing and display to dedicated viewers, making it difficult for non-dedicated client devices to efficiently handle and view DICOM images due to the need for complex algorithms, proprietary libraries, and inconsistent image compression techniques.
Innovation Solution
A method to normalize DICOM images by converting them into consumer-grade formats like JPEG-XR, reducing the complexity of image data and processing requirements, allowing for efficient transmission and display on a variety of client devices without the need for dedicated image processing resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If DICOM images are processed using the DICOM standard, then image quality and dynamic range are maintained, but device complexity and processing requirements increase significantly
Solution Approach 1:
The patent introduces a normalization server as an intermediary component between the DICOM image storage system and client devices. This server converts complex DICOM images into simplified normalized images with reduced pixel bit depth (e.g., from 16-bit to 8-bit), thereby reducing device complexity requirements while maintaining acceptable image quality for viewing purposes
Solution Approach 2:
The patent changes key parameters of the image data during normalization, specifically reducing the bit depth of pixel values from 16 bits to 8 bits, and converting from grayscale to RGB color space. These parameter changes simplify the image representation while preserving visual information, allowing standard client devices to process and display images without requiring specialized DICOM processing capabilities
2Loss of time
If compressed DICOM images are transmitted to reduce data transmission time, then transmission efficiency improves, but image manipulation consistency deteriorates
Solution Approach 1:
The patent applies normalization and compression operations in advance on the server side before transmission to client devices. By pre-processing images to a standardized format with reduced complexity, the system achieves efficient transmission while ensuring that all client devices receive consistently formatted images that can be manipulated uniformly, eliminating the need for each device to handle multiple compression formats
Solution Approach 2:
The patent transforms image parameters during the normalization process, converting from 16-bit grayscale to 8-bit RGB color space. This parameter transformation reduces data size for faster transmission while creating a standardized format that ensures consistent manipulation across different client devices, as the normalized format eliminates the variability introduced by different DICOM compression algorithms
3Adaptability or versatility
If 16-bit grayscale images are converted to 8-bit RGB color space, then device compatibility improves, but image data precision is reduced
Solution Approach 1:
The patent systematically changes image parameters by converting from 16-bit grayscale to 8-bit RGB color space. This transformation reduces the precision of individual pixel values (from 65,536 possible values to 256 per channel), but compensates by distributing information across three color channels, maintaining overall visual fidelity while dramatically improving compatibility with standard client devices that natively support 8-bit color formats
Data Source
AI summary
A method, apparatus and computer program product are provided to process a compressed image, such as a normalized DICOM image. A method may receive a compressed image having pixels that each have by a pixel value having a predefined number of bits in a grayscale format. The method may decompress the image by mapping the pixel value into two channels of a multi-channel output buffer, such as a red-green-blue (RGB) output buffer, such that each channel has fewer than the predefined number of bits. The method may render a scene of the image based upon camera coordinates and, for a rendered pixel, sample a plurality of RGB values, convert the plurality of RGB values to the grayscale format having the predefined number of bits, perform an interpolation of the RGB values following conversion to the grayscale format and perform a window level operation on the result of the interpolation.